Help with Phillips EV4437A PA AMP conversion > speaker impedance

[ulysses]

hey guys

im after some advice from someone with technical expertise in older OT's

ive got an old Phillips EV4437A PA amp that I want to convert to a bass amp. I'm thinking i'll use the preamp section of an ampeg SVT eventually. for now, im just running my guitar into the mic jack and it's sounding pretty good.

i am concerned about the mismatch of the impedance of the speaker. im running it into a 16ohm 12" inch speaker. it's working fine, but im sure it's wrong and dont want to damage the speaker or the OT.

here is the schematic of the EV4437A http://www.amplifiers-with-valves.nl/images/big/EV4437A-gr.pdf
and here is the schematic of the EV4437 (no A) http://www.amplifiers-with-valves.nl/images/big/EV4437S.pdf

if you notice the wiring to the speaker from the OT is different. in the EV4437A the speaker is connected to a 100V tap, where the EV4437 is connected to a separate tap all together.

in this document (http://dalmura.com.au/projects/Renovating%20PA%20amps.pdf) on page 15 it says "The common Philips EV4438
(A,B) PA amp has 50V, 70V and 100V winding taps (300Ω, 600Ω, 1200Ω), but also a separate 10V (12Ω)
feedback winding, based on a 9.1kΩ PP OT for 6GW8 / ECL86 valves." -- i suspect that the 2589 OT has similar specs (300Ω, 600Ω, 1200Ω).

is there a way i can test which suitable ohm speaker is required for the unused 3.8v tap on the 2589 OT? is that tap suitable at all?

or is there a way to bring down the 600Ω tap to 16Ω without using another transformer like this one -->http://www.hammondmfg.com/collins.htm

or does it really make no difference if i am plugging a 16Ω speaker into a 1200Ω tap?

cheers for reading!

[ulysses]

hmmm

maybe it has something to do with this?
http://www.horrorseek.com/home/halloween/wolfstone/Audio/spk70v_70VoltSpeakers.html

or this
http://www.ozvalveamps.org/optrans.htm (down the bottom of the page)

[Rob Strand]

My understanding of "constant voltage"  speaker outputs is it produces  xx Volts rms when the amplifier is full power.     Unless you have information stating otherwise, the relationship between xx Volts and impedance depends on the amplifier power rating.

P  =  Vrms ^ 2 / R

Vrms = sqrt(P/R)

R   = Vrms^2 / P

*** However ***

Apparently the output voltages on that schematic are wrong.
The numbers are impedances (25 ohm, 50 ohm, 100 ohm).

The trick to drive lower impedances is to drive between the output windings:
- between 100 and 50 ohm  taps for a 14.7ohm load.
- between 50 and 25 ohm taps for 7.2 ohms load.

Through the wonders of the internet it is all here:

http://www.guitargear.net.au/discussion/index.php?topic=38519.0

[A caveat with this method is it assumes output winding resistance is sufficiently low.]

[PRR]

Ulysses - go post on that other forum.

Without the beast in my hand, I would be 99% sure 100 50 25 are VOLTAGES. However trobbins over there says different, and I know he is no fool. As he has seen the beast and I have not, best listen to him.

> EV4437 is connected to a separate tap all together.

I believe this shows an internal 4-inch part-Watt monitor speaker. The Real Load will connect to 100V 50V 25V as appropriate.

> "EV4438 (A,B) PA amp has 50V, 70V and 100V winding taps (300O, 600O, 1200O)"

You can NOT take Volts and Ohms for one (low power) amp and apply them to another amp of different (14X greater!) power.

> a 16ohm 12" inch speaker

The inches hardly matters; except this *appears* to be a 120 Watt amplifier, and only a few Twelves will safely absorb that much power. (EVM 12x, JBL E-120)

Volts, Ohms, Watts.... remember your Ohm's Law.

At assumed 120 Watts, and 16 Ohms:

100V === 83 Ohms
50V ==== 21 Ohms
25V ==== 5.2 Ohms
10V(*) = 0.8 Ohms
3.8V(*)  0.12 Ohms
(*) these taps may not actually support full power output

The "50V" winding is best-fit for 16 Ohm load. Check: 50V squared is 2500, and 2500 divided by 16 is 156 Watts. In fact we expect the "50V" to sag to say 47V with the heavy load, and 138W is not that much more than rated power. Amps like these usually had "some" headroom for mis-calculated loads.

The 25V tap is a fair fit for either 4 or 8.

Hammond's "Collins" part is rated 12 Watts. You do not want to buy ten of these, even if they could be cobbled into a 10X part.

There *are* line-matching transformers to connect speaker impedances to voltage-rated outputs. But a 120W job is a lot of money, and weight; also "speaker impedance" is usually 4 or 8, while 16 is mostly out of style.

[ulysses]

thx for the replies guys

i have been through that thread on the other forum and had hoped to find some more experience here on diystomp.

i saw that someone thinks that they are incorrectly marked as voltages instead of ohms. i dont believe this is right. i have read numerous places on the web that they are in fact voltages for line driving in large pa applications ie, office block speakers on every floor etc.

when i plug it into a 16ohm speaker it seems to work fine. but for how long??

cheers

[ulysses]

i checked the OT, there are no wired soldered to the first two taps (where the EV4437 (not A) speaker is connected to) , and nothing soldered to the 50V or 25V taps.

cheers

[Rob Strand]

i saw that someone thinks that they are incorrectly marked as voltages instead of ohms.

To be honest, I'm not sure what to believe.  The interesting thing is the poster that indicated the error measured the output power and got a very sensible value.

Following from,

http://www.amplifiers-with-valves.nl/
Then,
http://www.amplifiers-with-valves.nl/images/LBH1015.pdf

This guy converted a similar amp and a he looks like he used the output taps the above poster used.   The schematic around the output stage has similar markings.

[PRR]

You can prove-out the markings with a sine generator, a modest power resistor, and an AC volt meter. .

A six 6CM5 amp is something over 100 Watts. At this power level, 100V and 100 Ohms are about the same thing. Put a 100r 10 Watt power resistor from 0 to 100.

Bring up a 200Hz sine until the output voltage is 10V. (Resistor should be a bit warm, not hot.)

Measure the voltage on the other taps.

If the numbers ARE voltage, they will be in-proportion: 5V, 2.5V.

If the numbers are Ohms, the voltages will NOT be in-proportion.

Report what you read.

If you copy-paste my tiny text you will see that I say what Rob said before he saw that other thread. I am still pretty-darn-confident these are Voltages, not Ohms. 25V and 100V make perfect sense in high-power many-speaker systems (schools, factories). 50V makes sense if a 10V line must be "balanced" to reduce leakage. However trobbins' post makes me wonder.

BTW: part of the confusion is that in North America we use 25V and 70V lines. I know 100V speaker lines are very common in Europe, but have never worked with one. You need some non-US/Can expertise. And the 25/50/100 combination could be unique to Australia. That's a nest of aussie amp geeks there, and IMHO an excellent source.

[PRR]

> 200Hz sine .... 10V

Contrariwise... with amp OFF you could connect a 6.3VAC winding through 10 or 22 Ohms to the 0 and 100 taps. (This OT will support many volts at 50Hz.) Read what you really get at the 100 tap (probably a hair high because your "6.3VAC" is un-loaded), then at the 50 and 25 taps. Report.

In any case: 16 Ohms on the "100" tap (either Volts or Ohms) is BIG mis-match, less power to the speaker and way excess waste heat in the tubes. I do not think this is correct or wise.

[ulysses]

thanks for the advice paul

i'll order a 100R 10W resistor on ebay and then report the results

cheers

[ulysses]

ah paul,

i didnt see your tiny text (wasn't sure what that was)

so if i connect it to the 50V tap, can i get rid of the 22nf and 470R components hanging off the 100V tap, or should i connect them to the 50V tap instead?

cheers

[ulysses]

ok so i hooked up the new speaker jack to 0V and 50V. i left the 22nf and 470R resistor hanging off the 100V to 0V.

interestingly the 100V line also has a connection to the power tubes, a wire not listed on the schematic.

anyway, it's sounding pretty bloody awesome. certainly a lot better than when it was connected to the 100V line.

im running it onto a 65W celestion creamback 12" speaker. i realise that i could blow the speaker if i run it too hard, but when cranked up to a similar volume as my bluesbreaker (30W OT) it sounds really nice.

thanks for all the advice guys

cheers

[PRR]

> a problem with hum at 50hz when i play a bass guitar note around 50hz.

https://www.dropbox.com/s/2k74qb4b8y3r961/modded-phillips-EV4437-to-ampeg-2017-06-02.jpg?dl=0

The oddity I noticed even before looking at a SB12: the cathodyne does not normally go to the NFB node. That does something odd which I'm too tired to think on. Since the "normal" connection is known to work, I would just do it normal and see what is different.

EDIT!-- I see this same connection on many recent re-drawings of the Philips Down Under products. I wonder if it is a too-clever trick, or a mis-draw?

Two different coupling cap values here is odd, and would make a difference on the lowest notes.

The 470K is not legal for *three* 6CM5. Book says three need 166K. You can cheat this a little, but not all the way to 470K; eventually there will be major trouble. Maybe buying Genuine Philips new in 1957, grid leakage would be low enough to get away with. Using old tubes, some made after the decay of the Golden Age, I would not push this spec.

https://s15.postimg.org/vtqzcpocb/Philips-_SB12.gif

> i suspect the filtering is not the best

You have about 3X the power output on 80% the voltage. Ampeg used 30uFd first cap. 30uFd*3/0.8= 111uFd to get about the same performance. 72uFd is significantly shy. My figures, ignoring what Ampeg did, say 220uFd is good. What did Philips put in it?

[anotherjim]

I'm pretty sure, with 100V line amplifiers, every speaker gets a matching step-down transformer in its enclosure.
http://www.canford.co.uk/100-VOLT-LINE-TRANSFORMERS
The link is an example, plenty more out there.
http://www.altronics.com.au/p/m1136a-10-60w-100v-line-pa-ei-core-transformer/

[PRR]

> 100V line amplifiers, every speaker gets a matching step-down transformer

This was worked out before. He has multiple taps. "The 25V tap is a fair fit for either 4 or 8."

[anotherjim]

Yeh, I know you found a work-around, but since I didn't see matching transformers mentioned, I thought I would. I mean, isn't that why it sounded crap with the speaker on the 100v tap?

[ulysses]

thx for your reply paul

i have updated the schematic and updated the amp. unfortunately it has not resolved the issue. in fact i t seems to have clarified the issue -- it seems to be some harmonic overtone distortion that i can now hear clearer than before.

when i checked the components by hand, i noticed that the schematic was wrong -- there were 2x 22nf and 2x 470k resistors, so there were not unbalanced caps leading to the power tubes.

i have uploaded the schem and an audio demo -- in case you have any further ideas you can hear the overtone distortion when i hold the notes.

https://www.dropbox.com/s/iwmis8qahk409bm/ampeg-sb12-workalike-audio.wav?dl=0
https://www.dropbox.com/s/vmj2p9d0e3qv1gl/modded-phillips-EV4437-to-ampeg-2017-07-18.jpg?dl=0

cheers